LiDAR, Point Clouds & Feature Extraction. A Beginners Guide

Shortages in Surveyors, reduced project times, and financial restraints are all having an impact on our industry. On the flip side of this ( or perhaps because of this ) is the progression in LiDAR, put simply - it's amazing. However, LiDAR on its own is not the solution. What turns un-intelligent points captured using LiDAR ( or photogrammetry ) into something truly useful is ...... feature extraction. Why?

Before covering the why, let’s delve a little deeper into just what each of the components within the process actually are using the descriptions provided by Wikipedia. I suggest it is a 3 part process.

Part 1 - LiDAR ?- ?Lidar?is an acronym for "light detection and ranging"?or "laser imaging, detection, and ranging".?It is sometimes called?3-D laser scanning, a special combination of?3-D scanning?and?laser scanning.

It is a method for determining?ranges?(variable distance) by targeting an object or a surface with a?laser?and measuring the time for the reflected light to return to the receiver. It can also be used to make digital?3-D representations?of areas on the earth's surface and ocean bottom by varying the wavelength of light. It has terrestrial, airborne, and mobile applications.

Part 2 - Point clouds - A?point cloud?is a set of data?points?in?space. The points may represent a?3D shape?or object. Each point?position?has its set of?Cartesian coordinates?(X, Y, Z).?Point clouds are generally produced by?3D scanners?or by?photogrammetry?software, which measures many points on the external surfaces of objects around them. As the output of 3D scanning processes, point clouds are used for many purposes, including creating 3D?CAD?models for manufactured parts, for?metrology?and quality inspection, and for a multitude of visualization, animation, rendering, and?mass customization?applications.

Part 3 - Feature Extraction – Feature extraction involves reducing the number of resources required to describe a large set of data. When performing an analysis of complex data one of the major problems stems from the number of variables involved. Analysis with a large number of variables generally requires a large amount of memory and computation power, also it may cause a?classification?algorithm to?overfit?to training samples and generalize poorly to new samples. Feature extraction is a general term for methods of constructing combinations of the variables to get around these problems while still describing the data with sufficient accuracy.? In short – defining attributes of data useful for analysis.

Advantages of LiDAR

1. Data is collected quickly ( and accurately ) - Laser scanning is the speediest land surveying method out there. Laser scanners can collect millions of data points in seconds, reducing the time and human resources needed to complete surveys. This method of surveying is perfect for projects that require a quick turnaround.?
2. Systems Can Be Mounted to Vehicles ( land or airborne ) - For projects that require information from vast landscapes or extensive networks of roads or railroads, laser scanning systems can be mounted onto land-based or aerial vehicles. The systems can then gather data from all of the necessary areas in a short space of time.?
3. Laser Scanning Can Reduce Health and Safety Risks - Using laser scanners to collect information from dangerous or hard-to-reach locations means teams don’t have to put themselves at risk. Instead, surveyors can stay safely out of harm’s way while the scanner does its job.?

LiDAR compliments traditional surveying and it is not designed to replace it.

for more on the advantages of LiDAR take a look at the article - https://blog.topodot.com/what-is-laser-scanning-and-how-can-it-be-used/

Point clouds - the benefits and limitations

A point within a point cloud is simply an x,y.z value in space, and many millions of these points are captured to obtain a representation of what exists. Great for capturing what exists and viewing ( given the right software ) but difficult to utilize for a number of reasons -

1. The more you zoom in the wider apart the points become within the viewing area, effectively reducing the resolution.
2. Point clouds are often extremely large and difficult to manage, access and display. If you have ever tried to store and access an average-sized point cloud ( say 10Gb ) then you will know how cumbersome they can be and how problematic they are within software not specifically designed for the purpose.??
3. Due to the nature of a point cloud, it cannot be used or readily used in its raw form within downstream applications such as BIM, GIS or CAD.?What is needed are the vector graphics ( points and lines with attributes ) to populate these downstream applications.

So why is Feature Extraction useful?

It is suggested that a thousand points do not tell me everything I need to know, ….. while about 15 extracted points tell me everything. In the image below, the point cloud on the left shows what it is ( a power pole ) but not what it contains or the details about it ( x,y, z coordinates, connection points, clearances, pole height, tilt etc ).

Given that the extracted features ( points, lines and attributes ) are vector graphics their size is a fraction of the same component within a point cloud. This means faster loading, easier manipulation, clearer display, and extended downstream application integration.

Feature extraction can also utilize imagery ( via calibrated images ) to assist with the process. No more guessing what that series of points/dots mean. As an example - details on a street sign can be clearly understood and used, which is not possible directly from the point cloud. For more on this take a look at the following article - https://blog.topodot.com/cad-software-turning-point-clouds-into-3d-models/

In conclusion - Point clouds are excellent at capturing what exists, but poor at telling you about the things you want to know. Of course, it always depends on your downstream needs but a small number of extracted features can tell you far more than tens of thousands of points in a point cloud. The ultimate goal is greater onsite safety for those capturing the survey data, improved productivity, and increased profitability for the overall business.

I agree this is the basics but it hopefully provides a starting point for understanding the process. Perhaps it's time to consider LiDAR and point cloud feature extraction for your digital surveying projects.